Sarpong et al. (2026) Creeping snow drought threatens Canada’s water supply

Identification

• Journal: Communications Earth & Environment
• Year: 2026
• Date: 2026-01-09
• Authors: Robert Sarpong, Ali Nazemi, Amir AghaKouchak
• DOI: 10.1038/s43247-025-03162-8

Research Groups

• Department of Building, Civil and Environmental Engineering, Concordia University, Montreal, QC, Canada
• Department of Civil & Environmental Engineering, University of California Irvine, Irvine, CA, USA
• United Nations University Institute for Water, Environment and Health, Richmond Hill, ON, Canada

Short Summary

This study introduces Snow Water Availability (SWA), a new metric quantifying water stored in snow-covered areas by normalizing Snow Water Equivalent (SWE) with snow cover fraction. It reveals that while overall SWA increased across Canada and Alaska from 2000-2019, significant losses in critical western Canadian headwaters and widespread insignificant decreases across southern Canada threaten water supply for a quarter of the country, impacting 86% of its population.

Objective

• To introduce and quantify Snow Water Availability (SWA) as a new diagnostic measure for snow water supply, accounting for sub-grid snow cover heterogeneity, and to analyze its spatio-temporal changes and associated uncertainties across Canada and Alaska.

Study Configuration

• Spatial Scale: Canada and Alaska, covering approximately 9% of the global land area, divided into 25 drainage regions and 18,060 grid cells. Data harmonized to a 25 km × 25 km grid system (NASA MEaSUREs FT-ESDR).
• Temporal Scale: 1 March 2000 to 31 December 2019 (19-year period), analyzed at monthly, seasonal (fall: October-December, winter: January-March, spring: April-June), and annual (October-June) timescales.

Methodology and Data

• Models used:
  • Conceptual snowpack model (within Canadian Meteorological Center's snow depth reanalysis)
  • Land-hydrology model (within ERA5-Land reanalysis)
• Data sources:
  • For SWA estimation:
    • Canadian Meteorological Center’s (CMC) snow depth reanalysis (24 km resolution)
    • CMC’s Snow Water Equivalent (SWE) reanalysis (for deriving snow density)
    • European Centre for Medium-Range Weather Forecasts (ECMWF) Land Reanalysis (ERA5-Land) for snow density and snow cover fraction (0.1° resolution)
    • Moderate Resolution Imaging Spectroradiometer (MODIS) global monthly Level-3 (L3) snow cover product (5 km resolution)
  • For benchmarking and analysis:
    • CanSWEv.7 dataset (in-situ snow depth and SWE)
    • Multi-Source Weather reanalysis data (MSWX) for temperature
    • Global Multi-resolution Terrain Elevation Data 2010 (GMTED2010) for elevation (1 km resolution)
    • Large Scale Climate Indices (LSCIs): Pacific Decadal Oscillation (PDO), Northern Oscillation Index (NOI), Pacific North American pattern (PNA), North Atlantic Oscillation (NAO), North Pacific Pattern (NPP), Western Hemisphere Warm Pool (WHWP), Atlantic Multidecadal Oscillation (AMO), Solar Flux, Western Pacific Index (WPI), Global Mean Surface Temperature (GMST), Pacific Warm Pool Area Average (PWPAA), and Arctic Sea Ice Area Average (ASIAA).
    • Statistics Canada water yield data.
    • Global Administrative Areas (GADM) shapefiles.

Main Results

• Snow Water Availability (SWA) was introduced as a new metric, defined as grid-averaged SWE normalized by snow cover fraction, to better account for sub-grid heterogeneity in snow water supply.
• The annual long-term mean SWA over Canada and Alaska was estimated at 996 ± 170 km³ during 2000–2019.
• Domain-wide annual average SWA increased from 799 ± 121 km³ in 2000–2009 to 1208 ± 231 km³ in 2010–2019, representing an overall gain of 91 ± 44% (Sen’s slope of 42 ± 14 km³/year) over the 19-year period.
• Despite overall gains, significant SWA losses (p-value ≤ 0.05) were observed in approximately 3% of the domain, primarily concentrated in the North American Cordillera (headwaters to major rivers in western Canada), mainly at elevations between 800–2200 m. This region experienced a substantial 71 ± 13% loss in SWA.
• The likelihood of significant SWA losses consistently increased with elevation, reaching 14 ± 2% in areas above 1500 m, while lowland areas (0–500 m) showed significant SWA gains (up to 40 ± 5%).
• Snow depth was identified as the key driver of significant changes in annual SWA, with a 99 ± 0.4% probability of alignment in negative hotspots and 99 ± 1% in positive hotspots.
• Global Mean Surface Temperature (GMST) emerged as the most influential Large Scale Climate Index (LSCI), showing significant dependence in approximately 50% of positive and 65% of negative SWA hotspots, indicating complex and regionally specific impacts of warming.
• SWA provides a more nuanced view of water availability than SWE, especially during the onset and offset of the snow season and in patchy snow covers, where SWE can produce false signals of water availability in snow-free areas.
• 14 out of 25 drainage regions in Canada and Alaska were identified as highly vulnerable, vulnerable, or at-risk to SWA loss, encompassing approximately 2.5 million km² (25% of Canada's total area) and potentially affecting 86% of the Canadian population and important ecosystems.
• Okanagan–Similkameen (DR3) and Assiniboine–Red (DR12) were classified as highly vulnerable regions, experiencing significant SWA losses with severe implications for agriculture and water supply.
• Uncertainty in SWA estimates, arising from the choice of snow cover and density datasets, can be substantial, particularly for trends, and exhibits spatial variability across longitudinal, latitudinal, and altitudinal gradients.

Contributions

• Introduction of Snow Water Availability (SWA) as a novel diagnostic measure that accounts for sub-grid snow cover heterogeneity, offering a more precise quantification of water stored in snow-covered areas compared to traditional Snow Water Equivalent (SWE).
• Development of a four-member ensemble for SWA estimation across Canada and Alaska, enabling quantification of trends, long-term means, and associated uncertainties at grid, seasonal, and drainage basin scales.
• Identification of specific "creeping snow drought" hotspots in western and southern Canada, particularly in critical headwater regions and populous areas, highlighting vulnerabilities to water supply for a significant portion of the Canadian population and ecosystems.
• Attribution of SWA changes primarily to snow depth variations and linking these changes to large-scale climate indices, especially Global Mean Surface Temperature (GMST), revealing complex regional responses to warming.
• Demonstration of the complementary nature and advantages of SWA over SWE, particularly in transitional periods and patchy snow cover conditions, for improved impact assessments and water management.

Funding

• Canada’s New Frontier Research Fund –Exploration (NFRFE-2020-01298)
• Natural Science and Engineering Council Discovery (RGPIN-2023-05760)

Citation

@article{Sarpong2026Creeping,
  author = {Sarpong, Robert and Nazemi, Ali and AghaKouchak, Amir},
  title = {Creeping snow drought threatens Canada’s water supply},
  journal = {Communications Earth & Environment},
  year = {2026},
  doi = {10.1038/s43247-025-03162-8},
  url = {https://doi.org/10.1038/s43247-025-03162-8}
}